However, this study evaluated fewer patients than 50 patients and some had bronchiectasis in addition to COPD. We extend the findings of this study to a much larger cohort of patients with prolonged follow-up. Our findings are also consistent with previous studies that have shown a positive relationship between serum MPO levels and poor cardiovascular health outcomes. For example, a previous large study has shown that serum MPO levels are associated with acute coronary syndromes among those who present to emergency departments with acute chest pain. Other studies have shown that serum MPO levels can predict incident risk of cardiovascular disease and death in healthy middle-aged and elderly subjects. We extend these previous observations by demonstrating that serum MPO levels are related to incident cardiovascular edna techniques efficient inventory monitoring programs sensitive species mortality in COPD patients who continue to smoke. While causality cannot be established on our study, these data nonetheless suggest that neutrophilic inflammation may be very important in the pathogenesis of COPD-related cardiovascular outcomes. It should be noted that the mechanisms underlying the relationship between MPO and cardiovascular outcomes may be different than that between MPO and FEV1 decline. There are several limitations to this study. First, the LHS cohort recruited only patients with mild or moderate disease. It is uncertain that these results can be generalized to COPD patients with more advanced disease. Second, we did not have multiple measurements of lung function after blood sample analysis, and potential factors such as the frequency of acute exacerbation contributing to rapid progression in COPD patients could not be explored in this study. Third, since only one venipuncture was carried out on the fifth annual visit during the study period, we could not evaluate intra-and inter-individual variation of serum MPO over time in patients. Moreover, we could not ascertain MPO levels in those who died or were lost to follow-up over the first five years of the study. As we did not have serial measurements of MPO over time, it is possible that some in quintile 5 became quintile 1 members and vice versa, resulting in misclassification of exposure status over time and thus diluting the relationship of serum MPO to FEV1 progression and mortality. Thus, our estimates may be conservative. Moreover, Agusti et al recently showed that 70% of patients with COPD who failed to demonstrate systemic inflammation at baseline remained “free” of systemic inflammation at 1 year follow-up, while 80% of patients with systemic inflammation at baseline continued to demonstrate an inflammatory phenotype at follow-up. Most importantly, patients with systemic inflammation had poor clinical outcomes including worse health status and increased risk of exacerbations. We extend the findings of this study showing that serum myeloperoxidase is a promising biomarker for tracking rate of decline in lung function over time and in risk-stratifying patients for cardiovascular mortality. Fourth, lung function measurements were made before and after venipuncture and as such the directionality of the relationship between serum MPO and decline of FEV1 is uncertain. It is possible that serum MPO “caused” a rapid decline but reverse causation was also possible. Furthermore, although serum MPO levels were significantly associated with accelerated decline in FEV1, the overall effect was very small and of questionable clinical significance. Fifth, although we assessed disease progression using FEV1 decline data, it should be noted that this end point is highly variable among patients with COPD with a low signal to noise ratio. As such, the use of other measurements would have strengthened the findings.